Nashtract

01 The dead end

The day rate
punishes efficiency.

Solving in three days what deserved nine means earning less for being better. Fixed price, in turn, pushes you to under-quote a problem whose depth you don't yet know; time-and-materials rewards slowness. Neither model truly aligns client and freelancer.

02 The protocol

Bound the risk,
share the deviation.

Nashtract shares the gap between estimated and actual effort by a rule frozen at milestone acceptance, and caps exposure on both sides — beyond the boundary, you renegotiate. Each party knows, in advance, their best and worst case.

P(T) = r [ M + β (T − M) ] M: estimated effort · T: actual effort · β: sharing factor frozen at the milestone · exposure boundary U on each side

The equation isn't there to impress. It makes the relationship deterministic.

03 Where it stands

An open exploration,
not a settled answer.

Honestly, I haven't fully cracked the mechanics of a truly win-win collaboration yet — but it keeps working on me. Nashtract is its formalisation: public, and open to challenge.

• Specificationv0.2 · deterministic core, no database or framework
• Traceabilityappend-only event ledger · attack suite
• Calibrationadaptive β (Bayesian) — under security review
• Licenceopen-source (MIT) · TypeScript

04 Backed by Y

The method, applied to
the relationship itself.